1 2 3 Previous Next

Business Discovery Blog

335 Posts
John Sands

Face the Facts

Posted by John Sands Sep 26, 2014

The human face is the ultimate visualization and we spend most of our life looking for answers in it. We often misread them and can actually be purposely mislead by them.

To help us get around this the lie detector was invented, the most common of those being the Polygraph which was invented in the 1920’s by John Augustus Larson, his parents were Swedish which just goes to show lots of innovation is born out of Sweden. His invention tried to get behind the face and find out whether people are telling the truth or not.

 

polygraph.jpg

 

The face and the brain are intrinsically linked but do we take everything on “face value”? We don’t and that’s because we can’t read a person’s mind and find out what is going on behind the expression.

We simplify them and send them in emails and texts because people can’t see our faces. I also listened to a very interesting program on the radio where due to the obvious limitations of this medium they were experimenting by using a swanee whistle to show an upward or downward trend on a line chart, try and do that with 100 data points.


Smile.jpg

 

Some people actually can’t interpret expressions due to a very sad condition that affects the brain and stops us from recognising faces it’s called Prosopagnosia and is caused by damage to a part of the brain called the fusiform gyrus. Every time they see a face they are unable to tell if they are a member of their family or a complete stranger. This underlines our reliance on viewing patterns from an upward trend on a line graph to the downward turn of a person’s eye brows.


As we look someone in the face we have to make a decision based on what we see, if they are well known we can pretty quickly judge their mood or reaction to something we have said, but with strangers that can be more difficult and often we can misjudge their moods or make the wrong assumptions based on previous experience.


Luckily with information we are able to read the mind of the visualization and look in to its brain. The tables, columns and rows of a databases mind are freely available and can help us make decisions based on facts, even the relationships and history can take us down the right path.

We are constantly making decisions based on information that is foreign to us and by just relying on data at face value we can be prone to misinterpretation. Look behind the smile you may be surprised.

John Sands

Blinded by Science?

Posted by John Sands Aug 14, 2014

With the advent of increasingly larger sources of data it is becoming even more difficult to view or imagine patterns within these data sources. This has become very important in areas such as science; the Genome project for example identified 2,000,000 Genes in the Human Genome, imagine looking at that as a series of numbers.

 

In 2013 Greg McInerny, Senior Research Fellow in Information Visualization for the Biological Sciences at Oxford University attempted to do some research on how visualization is used by scientists. There is an excellent blog on this published by @FutureEarth. Scientists are inherently skeptical of visualizations. Moritz Stefaner referred to it as “Dumb Blonde Syndrome” the idea that if something looks good, it is suspect. But even skeptical scientists are coming round to the idea that visualizations have their place in detecting patterns and outliers within massive amounts of data.

 

molecule.jpg

 

 

The visualization above shows the structure of a molecule. This would be impossible to view with the naked eye and can only be viewed by rendering a visualization utilising huge amounts of data. But its not just a case of taking huge amounts of data and creating a pretty picture, the following example proves the point.

 

piechart.jpg

 

 

 

It is impossible to view all of the slices and don’t even start to work out the percentages.


A good visualization becomes even more important when the stakes are really high. In the pharmaceutical industry it takes on average 12 years to take a drug from discovery to market and the process can cost around $4 billion. Only 10%-20% of new drugs make it to market and at any point the process can fail either due to adverse patient reactions or the drug just not being as effective as first thought.

You can imagine anything that can increase the likelihood of a drug getting to market is embraced. Data visualization can allow Researchers and Data Scientist’s to explore hugely complicated data sets and also then relate discoveries to non-technical audiences such as investors and regulators by using story telling.


What this says is that although we concentrate on specific subjects such as, Visualizations, collaboration, and storytelling none of these can work in isolation. The scientist will not trust the visualization without data and you can’t rely on data on its own without collaborating with your peers. So what you need is a harmonic join between the three factors.


Please don’t think I am trying to simplify things there are obviously many more pieces involved in this complex puzzle. But as the heat is turned up in the visualization arena and battle is joined between the main players, we will see the creativity of many a web developer let loose on even more and more fantastic visual delights. But embrace the scientist in you and look for substance in that style.


@QlikJohn

John Sands

Sensory Data

Posted by John Sands Jul 21, 2014

Why do people still go shopping rather than using the internet all the time? After all, the internet is normally quicker and your local shopping centre can never hope to have as many brands as you can find using your PC or tablet. It’s partially a social thing - we like to meet up with friends have something to eat and a chat. While we are shopping it’s a tactile thing - holding the item (or even smelling it!) as this reinforces the fact that it is real and of good quality. Retailers know this and try to lure us in with bright signage and interesting aromas. Once we have purchased our items we sit down with our friends over a drink and compare our purchases hoping for compliments or praise.


Ladies.jpg(Getty images)

 

When we are shopping on the internet we have none of this sensory feedback, apart from maybe a few flashing banners. We have to trust what the retailer has told us about the item and trust that what they say is true even down to the dimensions. This is difficult, and why some people will go to the shop see the product and then buy it on the internet.

If we are like this when shopping for clothes then consider how people  behave when looking at millions of pounds, euros or dollars of sales figures and inventory items.  Just like when internet shopping, we are compelled to trust the numbers and words we see on the screen even though there is much more at stake, the money is not in a safe somewhere or the inventory may be 1000’s of miles away. This is even more important when we talk of fully optimized supply chains where stock is kept to the bare minimum.

 

Logistics.jpg

(Getty images)

We solve this problem of trust by working in groups and comparing against known facts we’ve encountered before. How were the sales of this product last year? What are the sales of comparable products? We then show these figures to our peers and ask for comment, slowly turning over the figures in our head.

So replace the attractive signs with compelling visualisations, the aroma with data scent (bit of a stretch I know) and the chat with friends with real time collaboration.You then have a business discovery tool such as QlikView, which helps people feel, test and believe the data they’re seeing.

None of this is new.  It’s just taken a while for software providers to catch up with the human brain!

I don’t like clichés, but idioms help us to understand common conceptions (and misconceptions).  As with many sayings, the origins of the idiomatic expression “a picture is worth a thousand words” are obscure.  However, its meaning as it relates to the BI space is clear; it “aptly characterizes one of the main goals of visualization, namely making it possible to absorb large amounts of data quickly”.

 

The power of visualization is undeniable, and Qlik has always been in the vanguard of using human visual acuity to help us get value from data. I’ve personally written and presented on this for years*, and Qlik continues to build amazing visualization capabilities.

 

There’s an issue with visualization though – in that it’s only part of the picture!  Charts alone are arguably just better descriptors of data – great for information delivery – but are only partially valuable when it comes to analysis. To be of more analytic use, the visualizations must be fully interactive, paired with and powered by navigation capabilities.  This combination enables the free exploration of the data pictured, and the data not pictured but in the dataset.  (See Curt Monash’s blog ‘Visualization or navigation’ for an independent take on this.)

 

If visualization is not paired with navigation two things occur.

 

First, people have to create more and more visualizations to try and replicate an exploratory experience, i.e., to picture more perspectives on the data.  This is not optimal.  It takes time.  Further, while linked visualizations are useful, trying to compare across many different visuals is difficult for people. It’s not uncommon for organizations using visualization tools that lack full interactivity to end up with many hundreds of charts because of this need.  Chart bloat is a problem; reading a large amount of charts is as hard trying to assimilate the meaning of many words.  Academic institutions often limit the length of doctoral theses to 100,000 words.   Assuming the cliché’s broadly correct, if a picture = 1,000 words then just 100 charts = PhD thesis!

 

Second, and more seriously, if the focus is more about picturing data than navigating and exploring it, then there’s simply less business value to decision makers.  Any initial picture describes the state of data, providing an answer to the ‘what’s happened/happening?’ questions posed by organizations.  Answering the ‘why?’ questions that immediately follow demands rapid iterations through the data, re-picturing it if needed, or using alternative search and navigation experiences, so decisions can be made.

 

Ironically, I’ll rely on a visual from independent analyst Cindi Howson’s rigorous ‘2014 Successful BI Survey’ report as a proof point here:

 

Chart from 2014 Succesful BI Survey.jpg

Source:  Cindi Howson, ‘2014 Successful BI Survey’, February 2014, pp40.  Used with permission.

 

I’m not sure if this picture tells a thousand words, but it certainly speaks volumes (cliché!) about the relative business impact of combined visualization and navigation, given that the BI industry average for delivering ‘significant business impact’ is reported as just 28%.

 

 

 

*For subscribers my 2011 Gartner paper 'Who's Who in Interactive Visualization for Analysis and Dashboarding' may be of interest.

John Sands

Laughter the Best Medicine

Posted by John Sands Apr 23, 2014

I have spent the last couple of weeks traveling round Australia and South Africa talking about Natural Analytics and it is fascinating watching the way this subject resonates across the globe. Even though my brain has been turned to mush by the different time zones, as soon as I am on that stage the adrenaline kicks in and I come alive. This is the same adrenaline that would have given me the extra strength to have avoided being eaten 1000s of years ago. So we adapt our innate abilities to help us even when the original reason is largely not valid anymore.

 

When presenting I use humor and this is another example of how we use our bodies natural abilities. Laughter actually triggers the release of endorphins, the body’s natural feel-good chemicals. Endorphins promote an overall sense of well-being and can even temporarily relieve pain. So hopefully when you have watched me present you feel a little better afterwards.

 

Laughter.jpg

 

 

So as we go through our modern life our body is using innate abilities that evolved to help us survive in a very different environment? Well I am not so sure that is completely the truth. Yes of course we are not being hunted down by Sabre Toothed Tigers or having to fight the tribe next door for the best camp site, but we are still to a lesser extent fighting for survival! I admit the chances of you getting a spear in the back as you walk to work are remote but the chances of your company losing a deal to a competitor because you have poor visibility of the sales cycle is much more real.

 

These abilities are within us and have evolved through natural selection to help us. Why does a bubble chart work? It’s because we can easily spot outliers, things that are different and that’s because our visual acuity developed long before our ability to read and write. Which one of the graphics below is easier to interpret?

 

 

Table.png

Graph.png

 

They are both a series of lines and numbers but somehow the line graph is much easier to interpret and see patterns. A lot has been talked about Natural Analytics and how it helps us tap in to innate skills and resources, my journey across the globe has helped reinforce my belief in the common sense of it and how it just works. So come on the BI industry catch up or you may be the one that is eaten by that Tiger.

 

@QlikJohn

The Associative Experience is about the dynamic experience the user has with the BI application. It's about answering that next unanticipated question. It's about exploring data freely without any predefined paths.  It's about quickly finding new discoveries. It is also one of the many facets that makes QlikView unique. Yes, the Associative Experience could be imitated by other software and if so, I'm sure the work involved is not as inherit as it is in QlikView. With QlikView it is automatic. At previous companies I tried to imitate it and it took a lot of work and I still could not get it right. Oh and BTW, imitation is the sincerest form of flattery so thank you, I digress, let's continue.

 

golden.pngI close my eyes, my brain is an enormous database and without any visual stimulation to help me, I ask myself the question: What is my favorite apple?  Golden Delicious pops into my head. Now, even though I know the answer immediately, my mind has collected enough data over the course of my lifetime to process and come to this conclusion. My decision is based on a combination of senses such as sight, smell, taste and texture. Over time, I have satisfied each one of these senses by querying the various combinations of flavors, textures, varieties and colors until I found my favorite. In BI terms, these criteria can be seen as Dimensions – the textual and descriptive component used to find my favorite apple.

 

The mind is processing various bits of data naturally from its years of information gathering and its surrounding context. It is inclined to ask more and more questions until the user is satisfied that enough information is received in order to make the correct or desired decision.  Note that these questions are not predefined or prescribed however; they are freely formulated based on previous results.

 

This process is the basis of QlikView’s Associative Experience and part of what we refer to as Natural Analytics.

 

The Power of Green, White and Gray

 

I open my eyes…I now imagine I am able to visualize and interact with this data and its surrounding context in a single location, a QlikView application. I visualize the dimensions I associated with the apple: its varieties, colors, flavors and textures. Possibly, another category is available for comparison such as vegetables. Measures, the numerical component of the data, are introduced and automatically calculated and aggregated on the fly very quickly - displaying how many are grown or consumed in each region. I can further analyze this information using a variety of filters that show all related selections while still retaining the ones that are unrelated. At first it appears to be akin to a traditional BI dashboard, but with traditional BI a linear approach to analyze data is commonly used. For example, with traditional BI, once values are selected or filtered, the surrounding data and other context that either may be related or unrelated is lost; removing any possibility of making new discoveries, not the case with QlikView.

 

greenwhitegray.jpg

So, with QlikView how do I visualize and maintain the aforementioned associations similar to those that were previously formulated within my mind? The answer is QlikView’s power of green, white and gray. By starting anywhere in the application and simply selecting one or more list box values, all other visualizations, selections and aggregations dynamically update based off of that selection without losing surrounding context of the unselected data. Selected data is highlighted in green, related or associated data is highlighted in white and unrelated data is highlighted in gray. I can simply see all other surrounding dimensions and their related or unrelated values based off my initial selection. This allows me to ask that very important next question. Selecting yellow and crisp from the select boxes – not only shows me what fruits are yellow and crisp but also what vegetables are yellow and crisp too – the selections in white. I have made a new discovery.  I have found vegetables that might appeal to my texture and color preference. The power of green white and gray helps guide me to my respective decision as well as prompts me to ask the next question that possibly I did not anticipate – such as which yellow and crisp vegetables might also appeal to my taste.

 

QlikView delivers the world’s first associative experience. It manages associations among data sets at the engine level, not the application level and stores individual tables in its in-memory associative engine. Every data point in every field is associated with every other data point anywhere in the entire schema allowing users to quickly and easily explore data freely and answer that very important next question.

 

So there you have my perspective on the associative experience. Tell me what you think.

 

Regards,

 

Michael Tarallo
Senior Product Marketing Manager

Qlik

@mtarallo

John Sands

Touch and Go

Posted by John Sands Feb 21, 2014

My career before Qlik went down many paths. I served in the Royal Navy and after that an electrician, life guard, fax machine engineer, delivery driver for my friend’s bakery, software trainer, product management and finally in to Pre-Sales at Qlik! Two things propelled me in my career choice; either necessity (had to pay the bills) or noticing a new opportunity and going for it. Towards the latter half of my career path it was technology that attracted me.  I’m continually fascinated by how the world of technology is changing. We are very lucky in the technology sector that things change very quickly and you never get bored - there is always something new to learn.

As a consumer we wait for technologies to appear and then make a choice about whether to use them or not. As a technology vendor things are not that simple! Making the decision on what technology to develop - to put your money on - can be a gamble.  Let’s not forget Sony with Betamax and Philips with Laser Discs…

 

185px-Betamax_Tape_v2.jpg220px-LDDVDComparison-mod.png

 

My first experience with QlikView was version 5 and to be honest it was not the fastest tool at the time, because it was an in-memory product and memory was very expensive and you were lucky if you had 32 mb of RAM on your PC and a 266 mhz P2 processor. But the gamble paid off and now my laptop has 4 GB of RAM and with high end servers 2 terabytes of RAM is available.

So Qlik got it right that time. However, the trick is repeating it.  It’s not good enough just to be right once, you have to keep on doing it and in my opinion Qlik has.

One of bets we’re making is to use HTML5. The title of my article is touch and go, that’s because HTML5, really unleashes the potential of any web based software especially in the arena of touch activation. In my opinion in future all software will need to have the possibility to be interacted with on a touch device, whether that is a tablet or a smart phone (or whatever comes next). For those of you who have attended the Business Discovery World Tours you will already have seen the new functionality of QV Next and how it has been designed for touch first. At our most recent employee summit we were blown away by the functionality just around the corner.

 

Typewriter.png

 

 

So finally, we will be able to start moving away from the 150 year old technology of the QWERTY keyboard which after all was designed to stop a mechanical typewriter from jamming.

Henric Cronström

The Key to Heaven

Posted by Henric Cronström Feb 4, 2014

“To every man is given the key to the gates of heaven. The same key opens the gates of hell.”

                                                                    [As told to Richard Feynman by a Buddhist monk]

 

 

 

To the Buddhist monk, these words were a general guide to how to live your life.

 

To Richard Feynman, the words were about knowledge and science: He was convinced that Science, per se, is neither good nor bad. It is just a tool – a tool that can be used for both.

 

Key.jpg

 

When I see these words, I think of some of the functionality in QlikView: Functions and features that were introduced to solve problems that would be difficult or impossible to solve otherwise: Triggers, Actions, Dollar expansions, Set analysis, Alternate states, Show conditions, etc.

 

These features are all keys to heaven. Correctly used, they can enable you to build an application that calculates and compares immensely complex things, while still presenting the data in a way that a user can understand and investigate further.

 

However, the very same functions are also the key to making user-hostile and unmanageable applications, e.g. through:

  • Set analysis that hard-codes a selection – instead of letting the user select interactively.
  • Excessive use of Dollar expansions, Variables and Show conditions, which make the application difficult or impossible to manage.
  • Alternate states that are poorly labelled, so that the user gets confused about which selection really is applied.
  • Triggers with Actions or Macros that perform navigation or selections that really ought to be user-initiated and not automatic.

 

Enable the user!

 

The user will learn to interact with data, if you only let him. Most users have very intelligent questions and want to navigate in data, explore and discover things. Let them do this.

 

But if you instead obscure the QlikView logic by introducing too much additional logic using any of the above mentioned features, the user experience will be a very different one. Instead of an active, smart user, you will produce a passive user that doesn't understand how to use QlikView effectively and instead uses the application as a static report.

 

Some pieces of advice:

  • Navigation and selections should be left to the user. Don’t automate this. Let the user make the selections and interact with data.
  • Label fields and charts so that it is clear what they show.
  • Avoid hard-coding filters. For example, if you want one graph showing the numbers for 2014 and second graph for 2013, you should not create two separate graphs with the years hard-coded. You should instead use a Trellis chart with year as first dimension.
  • Avoid using Triggers and Macros.
  • Always ask yourself “How is this going to be managed? Is this a manageable solution?”

 

Don’t let the QlikView functions get in the way of making a user-friendly and manageable application. Instead, use them wisely.

 

Simplicity.

 

HIC

John Sands

Lost in Translation

Posted by John Sands Jan 14, 2014

Happy New Year everyone and as storms smash against my window and clouds just can’t seem to stop dumping water on my head I start thinking of travel and holidays (picture below shows how stormy it really is close to where I live).

BcLBvWXIQAAh1ar.jpg large.jpg

I live on the South Coast of England and I am lucky enough to be able to make frequent visits to Brittany in France with my wife. When we first started visiting our French was very basic. Whenever we wanted to order a meal we had to point at pictures and hope it was what we thought it was! This was ok to start with, the  pictures made our life easier when ordering a meal, but what if we wanted to find out more? Then it became clear that a nice picture had its limitations. So we went to lessons and learnt how to communicate in French. It’s not easy to learn a new language, but it’s very satisfying.  At last we can get behind the lovely picture of Croque Gagnet, and find out that it’s a toasted sandwich with Gouda Cheese and Andouille sausage!


Croque.jpg


So why the culinary tale of France? QlikView uses a scripting language to give access to some of the more powerful aspects in the platform. Some people feel that a scripting language is complicated and forms a barrier to the adoption of QlikView. “Au Contraire” I cry It is the power of the QlikView scripting language that leads us to more discoveries within our data. So let’s do some very simple scripting. I have a spreadsheet and it contains five pieces of information (fields), here is the script that QlikView uses to load that information.

 

LOAD Customer,
Zip,
City,
Address,
Customer_ID
FROM
[C:\Data\prescriber data.xls]
(
biff, embedded labels, table is Sheet1$);


So to translate, we are asking QlikView to find Customer, Zip, City, Address and Customer ID in the prescriber data spreadsheet. It’s just a matter of getting over that initial trepidation when learning something new.

  I don’t expect everyone to be expert writers of QlikView script, the same as I am not an expert of the French language, in fact “je ne suis pas très bon”. So to get started on that journey here is how we help you out, we have free training and a free download of QlikView so what could be simpler! Start the New Year by learning a new language and explore the power behind the pictures.

Nope, this isn’t a pastiche on Harry Potter, but a reflection on the reality on how far GUIs can take you.

 

Wizard.jpgNot long after I joined QlikTech, I attended a dinner with a group of QlikView users.  One of my dinner companions, who’d been using QlikView for a long time, had also been trying out a visualization tool.  He said something about his experience of using the two that struck me as very insightful: “What I can build in QlikView in 30 minutes, I can build in 10 minutes in a visualization tool.  What I can build in a day in QlikView, I cannot build in the visualization tool.”

 

Based on my interaction with customers and prospective customers in the last few months, it’s increasingly obvious that people are beginning to get it.  In their evaluations they are starting to see past the initial ‘wow!’ of an easily built chart, and comparing based on front-end experience and back-end power.  Without the back-end power of QlikView to speed iterative, unconstrained analysis any front end would quickly lose its sheen anyway – pretty charts count for little if performance is poor, as users won’t or can’t wait.   (QlikView’s strength here was borne out by BARC’s recent BI Survey, which ranked QlikView #1 for query performance vs. other ‘Visual Analysis & Data Discovery Vendors’).

 

What I still hear however is the comment (usually originating from our competitors) that ‘scripting is bad/outmoded/old-fashioned/too hard’.   Wrong, wrong, wrong.  Scripting is an aspect and evidence of the power of any platform.  QlikView has always been a Business Discovery platform, not a single use tool, and will continue to be so with QlikView.next.

 

Wizards are great – and QlikView has them too for data integration - but they’re an option for simpler scenarios.  It’s disingenuous to claim that a graphically driven wizard in any software product can facilitate building as broad a set of applications as a scripted development language can.   Any GUI or step-through wizard has to present a small set of options to the user – if not it becomes complex and confusing to use.  As such, at some point users inevitably come up against the edges of what they can do with a wizard: if they don’t, if all of the functionality of a BI product can be developed and deployed via a wizard, then that product must be limited in the value it can provide.  Again: “What I can build in a day in QlikView, I cannot build in the visualization tool.”

 

Finally, if users cannot build what they want in a simple BI tool due to its backend limitations, then they are going to have to do scripting anyway, just externally to the BI product probably using (yup – you guessed it) custom SQL code, and that comes with its own set of demands and limitations.  Just try writing the SQL to handle OUTER JOINS between multiple tables to display in a viz tool vs. the script to do so in QlikView (LOAD * FROM…).  I know which is simpler - but again that speaks to the power of QlikView’s backend – its associative engine.

 

For QlikView the answer is wizards and scripts, to access the power of the backend and deliver the widest value via the most appropriate interaction for each type of user.

John Sands

A Time of Growth

Posted by John Sands Nov 26, 2013

The title of this blog is not just about economic growth but also the growth of my mustache in support of the Movember charity http://mobro.co/johnsands64! My wife will just be glad when it’s gone but I am getting rather attached to it.

IMG_0887.JPG.jpg

OK, let’s get back on track and try to remove that disturbing image from our memory. I have just finished a tour of four countries presenting at the QlikTech Business Discovery World Tours.  The overall feedback has been really good. I have chatted to many people and one comment really stood out, and it was not about QV Next or Natural Analytics, but rather it was about the presenters. They said you all seem to have a real excitement and energy when talking about QlikTech – which is great to hear.

In my last blog I talked about it not just being features but also being the ethos of the company that make Qliktech special and this comment really underlined that. I presented in Spain and Portugal and even in a time of economic downturn we had nearly 700 attendees in the two countries. 

 

What makes so many people gather in one place and take time out of their busy days? Is it the offer of free food and drinks? Or that feeling of not wanting to miss out? I think it’s the latter. It is our natural curiosity and wanting to be part of something special and the potential that it might just make us a star.

 

A lot has been said about Natural Analytics and how it uses our innate human behaviors such as association, comparison, anticipation and narration. This is nothing new and those same natural behaviors drive us to seek out new things and want to be part of a movement that is disrupting the norm. Wouldn’t it be great to have the inside track on the next big thing?

 

Over the coming months there is going to be a lot of exciting things happening at QlikTech and the World Tours are just the start. Back in 1993 we started a journey driven by disruption. We are not about to stop that disruption in fact we are doing it again, based on the novel concept of creating a software solution that works the way we do rather than the other way round.

 

So trust your natural instincts, and I promise not to show pictures of me with a mustache again.


Managing a Business Intelligence (BI) deployment of any size is not trivial. Hardware, software, data and analytical applications, all require individual attention to ensure their stability and effectiveness. In order to ease maintenance, increase productivity and prevent an ‘out of control’ BI deployment, it is essential to implement standards and best practices as the groundwork for application governance.


The QlikView Deployment Framework (QDF) introduces a set of standards and best practices to QlikView. It’s a result of the combined experiences of our customers, partners and QlikView experts. Its main purpose is to address structure, organization and resource reusability within a QlikView environment, in turn reducing development time and increasing QlikView manageability.


So, what is a Framework anyway?

framework.png

 

In its simplest form, a Framework can be described as an ideal structure of something. Something that can also be designed to reduce the most common and repetitive tasks. Let's take a moment and think of your home. Before it can provide shelter, entertainment or even privacy, you need a basic structure or frame (ah, ha) that will support … well, everything else. Without a frame, your home would be unprotected from the weather, there’d be no place to plug in your PlayStation, and you could not shut out your mooching neighbor. The frame enables you to develop your home into something suitable and of value. If your home’s design is found effective and provides the most common components and amenities desired by the mainstream population – it may become a model for building the perfect home. Furthermore, prefabricating some of the home’s most common components will make it even quicker and easier to assemble (think of pre-assembled trusses), in turn saving time and money. In comparison, this simple analogy describes the basic concepts behind a software framework.


Ah, I see, so, then what is a Software Framework?


mix_apps.pngThe objective of a software framework is to make it faster and easier for developers to build and maintain applications. Frameworks typically ‘bundle’ together a collection of components in a simple to use form. Notable software frameworks include Ruby on Rails (Web), .NET Framework (Microsoft), Prototype (JavaScript), Spring (Java) and the various iOS Frameworks (Apple). For example, iOS developers may use the iOS Bluetooth Framework to accelerate the development of the Bluetooth portions of their project.

 

The QlikView Deployment Framework builds upon the same light-weight application development model that makes creating QlikView analytics easy. Assembled with a set of standards, modules and QlikView tools, QDF introduces consistency, reuse and increased control across all QlikView applications.

 

The following is a list of standards and practices defined by QDF:

 

Resource Container Architecture – A core file system structure composed of directory folders (referred to as ‘containers’) bound to the framework. It organizes, secures and stores projects and various objects used by QlikView applications and other QlikView deployments. By using QDF, all different document types and functions have their own respective place in their own folder structure and can be moved easily without effecting the operation of the QlikView application(s).

 

Container Map – An internal structure which maintains links between containers in order for objects (scripts, language settings, color schemes, variables, expressions, data connections, etc.) to be shared and reused by QlikView applications. The container map is managed with the Container Map Editor.

 

Container Naming Convention – A set of standard and recommended unique names used to name the file system folders according to application, department, project or resource. These can be modified as needed without affecting the functionality of the application.

 

Centralized Variables – A repository that stores system, user expressions, and other variables as user defined names. These names are referenced in QlikView projects as variables or can be called with the QDF sub-functions. Variables can be local to a project or global to the entire deployment. They are created and maintained with the Variable Editor.

 

Initiation Script – A pre-defined QlikView Load Script which initiates QDF, common modules and sets variables used within all QlikView applications that are bound to the framework.

 

Sub-Function Library – A set of modules used by QDF that contain callable sub-functions and practical examples. These can be easily included in QlikView applications. Examples include sub-functions for advanced calendars, document and data migration, data parsing, data exporting and linking resource containers and variables.

 

qdf_diagram.png

Fig. 1 – Simple resource container logical diagram (left), physical file system structures (right)

 

What value will QDF bring?

 

There isn't always ‘one’ answer for best practices, but rather a collection of proven results from customers, partners and QlikView experts that provide enterprise scale, efficiency and governance for QlikView implementations. The QlikView Deployment Framework is the result of this collaboration. The value an organization realizes from the QlikView Deployment Framework will vary depending on the size, complexity and customizations of the deployment and its applications. Large and Enterprise QlikView deployments will greatly benefit from QDF due to ever-changing business requirements, data additions and organizational growth. It is recommended that smaller deployments investigate QDF but it may not be necessary to implement. Generally speaking, QDF will shorten development time for each new QlikView deployment, allow easier maintenance and reuse of resources and provide an overall application governance model.

 

How can I get started with QDF?

 

The QlikView Deployment Framework is available to members of the QlikCommunity under the QlikView Deployment Framework Group. Members may start a discussion, report a problem and receive the latest updates on the QDF software and documentation. After membership is granted, navigate to the Content tab to download the QDF Deploy Tool and review the QDF documentation to get started.


Join now: http://community.qlik.com/groups/qlikview-deployment-framework


Please review this Technical Brief (members only) for a deeper introduction on QDF


Regards,


Michael Tarallo

Senior Product Technical Manager

QlikView and QlikView Expressor

Follow me on Twitter @mtarallo

As a kid I loved to play video games like Defender, Battlezone and Elite.  The games themselves had a lot in common – they all had blocky 8-bit graphics, involved shooting stuff, made loud explosions and used ishiness

 

Defender caption 2.jpgIshiness? 

 

 

Coined by QlikTech’s John Teichman, ishiness means having a quality that gives people the ability to maintain an overall sense of a data set and where they are within it.  (According to John the term’s a corruption of the way we use ‘ish’ as a suffix in English to denote that something’s broadly right.)   Whatever it’s called it’s a fantastically helpful when looking at large or complex datasets.

 

Below are two examples of ishiness in the pre-release version of QlikView.Next (and of course maybe subject to change as such).  The first is a video of a column chart with a grey ‘ishy’ window below it, showing which part of the overall data appears in the main visual.   It’s a simple idea, but very effective, particularly when paired with interactive selection methods, like those shown in the second half of the video.

 

The way it works reminds me of the side-scrolling maps in old video games, which gave orientation and situational awareness clues to gamers.  To put my old Gartner hat back on for a minute, I suppose you could say that ishiness is one aspect of the ongoing gamification of analytics.

 

The second example video shows what happens when zooming and panning a scatter plot.  Note the small dots that appear and move around the borders of the chart.  Without these small ishiness dots which appear around the main visual it would be all too easy for people to forget where they are in the data overall.  In other words to lose the ‘information scent’ they were following and most likely the insight they’d make.

.

QlikTech’s been thinking about this need for orientation within data for a while now, and it could be argued that the mini charts that can be put into tables in QlikView11 show some ishiness too

 

Capabilities like this are important - as the datasets people want to assimilate and analyse get ever bigger the analytic software we use has to help us navigate and orient ourselves in large information spaces easily and quickly.  Ishiness is one aspect of the natural analytic capabilities that do exactly that.

 

Oh yes, and it happens to be pretty cool too.  That’s ishy, man.

According to Gartner both CIOs and CFOs list BI or Analytics as their top priority.  Despite this focus, organizations have had a difficult time deploying BI successfully. According to the BI Scorecard, BI adoption rates have essentially hovered around the 25% mark since they started surveying companies back in 2005.   (That was the year that the first YouTube video was uploaded.  And, the iPhone wasn’t even introduced until 2007!)

untitled.png

This discussion is usually phrased as driving a more “analytical culture”: one where there is a general desire and willingness to make decisions based off of factual quantifiable data versus intuition.  Usually, I would say that culture is one of the hardest things to change.  But this time the likes of Google and Apple have done much of the heavy lifting for us.  Rather than rely on gut feel, many of us have become information junkies over this same time period.  So why is BI adoption so hard?

 

Multiple issues are at play of course and a successful BI strategy will need to address each of these.  Here are my top three factors contributing to lackluster BI adoption.  What would you add to the list?

 

IT Centric vs. User Centric Models of Delivery

 

A typical BI report, visualization or view can answer a question which is well understood in advance.  However, when business users have a follow-up question (which they will), they typically have to go back to IT.  In this IT centric model of BI delivery, users are unable to answer business questions in a timely manner so they become frustrated with the system and with IT.

 

In a user-centric delivery model, IT is still hugely important.  However, instead of asking “What question does my business want to answer?”, we instead focus on “What types of questions does my business want to explore?”.  In other words, IT does not focus on delivering a solution but instead a tool.

 

Lack of Business Partnership

 

According to a recent study that looks at the changing role of the CIO, one striking finding is that only about half of CEOs felt that CIOs understood their business and the problems that face them.  Long gone are the days when monolithic technology solutions are seen as long term investments.  On the other hand, evaluating and selecting tools department-by-department doesn’t leverage the economies of scale that large enterprises enjoy.


IT can still invest in stable centralized technology.  But rather than a one-size-fits all approach, IT needs to deliver flexible tools at the enterprise level that provide services which can be rapidly adapted to the changing needs of each individual line of business. In this way, IT can deliver stability as well as being seen as responsive.


Innumeracy

If we want to grow BI beyond just a deployment of tools for our power users, we need to consider the skills of our entire user community.  In the U.S. where I live, this last point recently became painfully highlighted.  A study published by the Organization for Economic Cooperation and Development places the United States as #17 out of 19 for numeracy skills.


As we race forward into what many refer to as an ‘information economy’ the ability to understand data and to think in numbers is becoming even more critical.  Training our employees on the data and the use of data is just as important (if not more important) than on how to use the tools themselves.

 

What would you add to the list? 

What issues do you think we need to address in order to improve our ability to drive pervasive business intelligence?

I recently came across an interesting blog post by Curt Monash entitled “Visualization or Navigation.”  In this post Curt proposes that ‘navigation’ is more important than visualization, and that this capability is a product of the platform technology that underlies a BI product.  A quote from the article..

 

“It might seem that a lot of the action in business intelligence revolves around ever-better visualization… But I don’t think that’s exactly right — rather, I see navigation as being a much bigger deal. And unlike most pure visualization, navigation usually depends strongly on underlying platform capabilities.”

 

We at Qlik couldn’t agree more.  That’s not to say data visualization isn’t important.  In fact, we have invested heavily in our visualization capabilities in QlikView.Next.  It’s just that visualization is only one piece of a broader set of capabilities that enable discovery and insight.

 

What Curt calls navigation we tend to think of as interactivity.  This means giving the user a natural and effective way to go beyond the information they are offered in the initial set of charts and graphs.  No matter how good a visualization is, it can’t possibly tell the entire story.  And furthermore, a good visualization will prompt more questions than it will answer.  That’s why we see visualization as a starting point, not the whole picture.

 

Another recent trend has been around the concept of data storytelling.  Several vendors including ourselves are moving in this direction.  We see data storytelling as a way to collaborate and persuade with visuals and narrative presented in a manner that articulates insights and opinions.

 

Storytelling.jpg

 

Again, interactivity is a key ingredient.  It’s not enough to publish static data stories that are still a form of one-way communication.  There is a need for interactivity and discovery even when telling stories.  The experience should be a dialog, not a monologue.  Data stories should allow for a user to seamlessly move between live, interactive analysis and presentation of pre-determined insights.

 

In the end, it all goes back to the strength of the platform.  Interactivity is only possible when the underlying technology supports it in a powerful and unique way.  QlikView’s Natural Analytics@ technology drives an interactive experience that is hard to match.

Filter Blog

By date:
By tag: